From a database of 133 patients (98 children and 35 adults) who underwent multiple blood sampling for ⁹⁹Tc^m-mercaptoacetyltriglycine (MAG3) clearance, we determined simplified algorithms allowing the estimation of clearance. A one-compartment model with two blood samples was applied. The best choices for the adult population were the 12 and 90 min blood samples, giving a standard error of the estimate (S.E.E.) of less than 10 ml min^-1 1.73 m^-2; for the children, the 10 and 80 min blood samples gave a S.E.E. of 20 ml min^-1 1.73 m^-2; for both the adults and the children, the 10 and 70 min blood samples gave the best results with, however, a S.E.E. of 19 ml min^-1 1.73 m^-2. The use of such a combined algorithm will therefore result in a degradation of the results in adults, suggesting that a separate algorithm for each group is preferable. We compared the accuracy of the two blood sample method to the one blood sample method based on previously published algorithms for children and adults, respectively. The S.F.E. was significantly lower, in adults as well as in children, using the empirical two blood sample method. This two blood sample method seems potentially useful for routine practice in adult patients. The advantages of using such a method in children is balanced by the practical problems inherent in the need to take a second blood sample during the first 10 min, at a time when the plasma activity is rapidly decreasing.